Computer control of braking systems for freight trains
Contents
Project
description
Background
Computers are ubiquitous in society today and
they have also entered the area of train control. Since decades computers have
been used in railway interlocking systems and in dedicated applications in the
trains themselves. Today networking computers are coming into use where
applications in the wagons of a train are interacting with each other and where
a train application is spread over all the wagons. An example is the braking
system where a computer system can give shorter response time and better means
to control the braking than the pneumatic system of today. Each wagon has its
own computer system controlling the brakes according to commands from the
locomotive. Such a computer system is a distributed real-time system.
Future computerised control systems will to a large extent be constructed as
distributed real-time systems. Data collection and processing can done near the
sensors and actuators and electrical or mechanical interface will be replaced
by data communication. This adds flexibility and safety to the controlled
system. The technology behind dependable, distributed real-time systems is
still in its infancy and although a distributed architecture is used for most
control systems today there are many question still
waiting for their answers.
One major question is how to achieve dependability with contemporary
technology. The techniques behind most of the data communication products on
the market today are not sufficient for safety-critical application. They offer
good solutions for most automation applications, but they fall short when the
demands on dependability are high. There is a need to gain more knowledge on
how to use today’s technology in tomorrows
safety-critical applications. There is also a need to gain knowledge on which
requirements must be put on computerised systems in trains with regard to the
needs of the users and what requirements can realistically be satisfied with
regard to technology. We have to push technology so that we can build and
verify safe systems that satisfy the requirements.
Strategy
We will study existing and proposed
applications of distributed real-time computers in trains. We will also study
the technology behind existing products and analyse their strengths and
weaknesses. Such studies will be a good starting point for the project and will
give all involved parties a good understanding of the application area - domain
- and the state of the art.
Further we will study the requirements on dependability of computerised control
systems in the domain. Dependability here includes availability, reliability,
safety and security. While modern computer systems generally satisfy the first
two aspects, they to a large extent lack safety and security. Safety will be an
essential property in tomorrows computer systems, if
they shall replace proven mechanical systems. Security will also be an
essential property, when computer system shall communicate in a wide area
network and intruders can have the possibility to get access to the network.
The studies will a basis for models of distributed train systems where the
dependability requirements can be validated, i. e. be
shown to represent the real needs of users. The models shall also be used for
verification of the techniques to construct dependable computer systems, i. e. to show that a particular technique can be used to
satisfy the requirements.
The domain knowledge and the models shall be used to find and evaluate
techniques to build safety-critical applications on top of existing technology.
Project members
- Jan Torin, PhD, Professor
- Håkan
Edler, MS, project leader
- Roger
Johansson, Lic. Tech
Publications
Dependability characteristics and
safety criteria for an embedded distributed brake control system in railway
freight trains.
Roger Johansson, 2001.
Report No 8, ISSN 1404-5001.
On
calculating guaranteed message response times on the SAE J1939 bus.
Roger Johansson, Jan Torin, 2002
Report No 10, ISSN 1404-5001.
On Communication Requirements for Control-by-Wire Applications
Roger Johansson,
Per Johannessen, Kristina Forsberg, Håkan Sivencrona, Jan Torin, 2003
In Conference Proceedings of : The 21st
International System Safety Conference 2003 (ISSC21) August
4-8 2003, Ottawa, Canada
A fault tolerant architecture for brake-by-wire in railway cars
Roger Johansson, 2003.
Report No 15, ISSN 1404-5001.
GAST - General Application Development
Boards for Safety Critical Time-triggered Systems
Roger Johansson,
Per Johannessen, 2004.
In Conference Proceedings of: The 22st International System Safety Conference 2004 (ISSC22) August 2-6 2004, Providence, USA.
A fault
tolerant architecture for computer based railway vehicle brake system.
Roger Johansson, 2004.
In press: Journal of Rail and Rapid
Transit, Proceedings of the Institution of Mechanical Engineers Part F,
Professional Engineering Publishing, Suffolk, UK.
Time and event triggered communication
scheduling for automotive applications
Roger Johansson, 2004.
Report No 17, ISSN 1404-5001, September 2004.
Collaborations
SAB WABCO/Cardo Rail AB
Funding
This work is funded by Charmec
Railway Mechanics - a VINNOVA Competence Centre
Programme Area 4, SD3 (System för
övervakning och drift)